Jm. Tahir et al., CONCURRENT ERROR-DETECTION IN FAST FERMAT NUMBER TRANSFORM NETWORKS, Computer systems science and engineering, 12(3), 1997, pp. 221-226
Citations number
12
Categorie Soggetti
System Science","Computer Application, Chemistry & Engineering","Computer Sciences, Special Topics","Computer Science Theory & Methods","Computer Science Hardware & Architecture
For many real-time and scientific applications, it is desirable to per
form signal and image processing algorithms by means of special hardwa
re at very high speed. With the advent of VLSI technology, large colle
ctions of processing elements can be used to achieve high-speed comput
ations. In such designs, some level of fault tolerance must be obtaine
d to ensure the validity of the results. Fermat number transforms (FNT
s) are attractive for the implementation of digital convolution becaus
e the computations are carried out in modular arithmetic which offers
three advantages: no roundoff error, no multiplications in the transfo
rm, and decomposition into fast algorithm analogous to the FFT. In thi
s paper we present a fault detectable array architecture for the fast
implementation of Fermat number transform. The results show that the d
esign offer concurrent error detection (CED) using very low hardware a
nd time overheads.